Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input

Estimating the hand-object meshes and poses is a challenging computer vision problem with many practical applications. In this paper, we introduce a simple yet efficient hand-object reconstruction algorithm. To this end, we exploit the fact that both the poses and the meshes are graphs-based represe...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Murad Almadani, Ahmed Elhayek, Jameel Malik, Didier Stricker
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Hand pose estimation hand shape estimation hand-object interaction graph convolution machine learning Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/cbcfc2045008438f8f0121b7a815ca3c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:cbcfc2045008438f8f0121b7a815ca3c
record_format	dspace
spelling	oai:doaj.org-article:cbcfc2045008438f8f0121b7a815ca3c2021-11-09T00:03:40ZGraph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input2169-353610.1109/ACCESS.2021.3117473https://doaj.org/article/cbcfc2045008438f8f0121b7a815ca3c2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9557272/https://doaj.org/toc/2169-3536Estimating the hand-object meshes and poses is a challenging computer vision problem with many practical applications. In this paper, we introduce a simple yet efficient hand-object reconstruction algorithm. To this end, we exploit the fact that both the poses and the meshes are graphs-based representations of the hand-object with different levels of details. This allows taking advantage of the powerful Graph Convolution networks (GCNs) to build a <italic>coarse-to-fine</italic> Graph-based hand-object reconstruction algorithm. Thus, we start by estimating a coarse graph that represents the 2D hand-object poses. Then, more details (e.g. third dimension and mesh vertices) are gradually added to the graph until it represents the dense 3D hand-object meshes. This paper also explores the problem of representing the RGBD input in different modalities (e.g. voxelized RGBD). Hence, we adopted a multi-modal representation of the input by combining 3D representation (i.e. voxelized RGBD) and 2D representation (i.e. RGB only). We include intensive experimental evaluations that measure the ability of our simple algorithm to achieve state-of-the-art accuracy on the most challenging datasets (i.e. HO-3D and FPHAB).Murad AlmadaniAhmed ElhayekJameel MalikDidier StrickerIEEEarticleHand pose estimationhand shape estimationhand-object interactiongraph convolutionmachine learningElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 136438-136447 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Hand pose estimation hand shape estimation hand-object interaction graph convolution machine learning Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Hand pose estimation hand shape estimation hand-object interaction graph convolution machine learning Electrical engineering. Electronics. Nuclear engineering TK1-9971 Murad Almadani Ahmed Elhayek Jameel Malik Didier Stricker Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input
description	Estimating the hand-object meshes and poses is a challenging computer vision problem with many practical applications. In this paper, we introduce a simple yet efficient hand-object reconstruction algorithm. To this end, we exploit the fact that both the poses and the meshes are graphs-based representations of the hand-object with different levels of details. This allows taking advantage of the powerful Graph Convolution networks (GCNs) to build a <italic>coarse-to-fine</italic> Graph-based hand-object reconstruction algorithm. Thus, we start by estimating a coarse graph that represents the 2D hand-object poses. Then, more details (e.g. third dimension and mesh vertices) are gradually added to the graph until it represents the dense 3D hand-object meshes. This paper also explores the problem of representing the RGBD input in different modalities (e.g. voxelized RGBD). Hence, we adopted a multi-modal representation of the input by combining 3D representation (i.e. voxelized RGBD) and 2D representation (i.e. RGB only). We include intensive experimental evaluations that measure the ability of our simple algorithm to achieve state-of-the-art accuracy on the most challenging datasets (i.e. HO-3D and FPHAB).
format	article
author	Murad Almadani Ahmed Elhayek Jameel Malik Didier Stricker
author_facet	Murad Almadani Ahmed Elhayek Jameel Malik Didier Stricker
author_sort	Murad Almadani
title	Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input
title_short	Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input
title_full	Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input
title_fullStr	Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input
title_full_unstemmed	Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input
title_sort	graph-based hand-object meshes and poses reconstruction with multi-modal input
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/cbcfc2045008438f8f0121b7a815ca3c
work_keys_str_mv	AT muradalmadani graphbasedhandobjectmeshesandposesreconstructionwithmultimodalinput AT ahmedelhayek graphbasedhandobjectmeshesandposesreconstructionwithmultimodalinput AT jameelmalik graphbasedhandobjectmeshesandposesreconstructionwithmultimodalinput AT didierstricker graphbasedhandobjectmeshesandposesreconstructionwithmultimodalinput
_version_	1718441424165273600

Graph-Based Hand-Object Meshes and Poses Reconstruction With Multi-Modal Input

Ejemplares similares